schottky diode question

T

tempus fugit

Guest
Hey all;

I'm planning on using a MAX4622 CMOS switch in a circuit I'm building. On
p.7 of the data sheet
(http://datasheets.maxim-ic.com/en/ds/MAX4621-MAX4623.pdf), it says:

"If power-supply sequencing is not possible, add two small signal diodes in
series with the supply pins and a Schottky diode between V+ and VL"

I've read a bit about Schottky diodes, and I see that there are several
different types (like with normal diodes I suppose). How do I know what I
should use for this purpose? Can someone recommend one?

Thanks
 
"tempus fugit" <toccata@quitspammingme.ciaccess.com> wrote in message
news:a5373$4a3030b3$d1d89dfd$29470@PRIMUS.CA...
Hey all;

I'm planning on using a MAX4622 CMOS switch in a circuit I'm building. On
p.7 of the data sheet
(http://datasheets.maxim-ic.com/en/ds/MAX4621-MAX4623.pdf), it says:

"If power-supply sequencing is not possible, add two small signal diodes
in
series with the supply pins and a Schottky diode between V+ and VL"

I've read a bit about Schottky diodes, and I see that there are several
different types (like with normal diodes I suppose). How do I know what I
should use for this purpose? Can someone recommend one?

Thanks

Sorry, I forgot one other question. The datasheet says to sequence V+, then
Vl, V-, then logic ins. If the V+ was 9v, and Vl was 5v, and I had a power
supply that went:

+15vDC =>7809 => 7805, would that be power supply sequencing?

Thanks again
 
On Wed, 10 Jun 2009 18:18:39 -0400, "tempus fugit"
<toccata@quitspammingme.ciaccess.com> wrote:

Hey all;

I'm planning on using a MAX4622 CMOS switch in a circuit I'm building. On
p.7 of the data sheet
(http://datasheets.maxim-ic.com/en/ds/MAX4621-MAX4623.pdf), it says:

"If power-supply sequencing is not possible, add two small signal diodes in
series with the supply pins and a Schottky diode between V+ and VL"

I've read a bit about Schottky diodes, and I see that there are several
different types (like with normal diodes I suppose). How do I know what I
should use for this purpose? Can someone recommend one?

Thanks
The need for extra diodes suggests that this part has SCR latchup
problems. If it does, other things, like signal spikes, could also
latch it up... and then it will fry.

Try to find somebody else's part. Maxim tends to be bad news, in
general, anyhow.

Why do you need a 5 ohm analog switch?

John
 
"John Larkin" <jjlarkin@highNOTlandTHIStechnologyPART.com> wrote in message
news:fpc035lsgurn8km6kr5otu2cg806lgf71a@4ax.com...
On Wed, 10 Jun 2009 18:18:39 -0400, "tempus fugit"
toccata@quitspammingme.ciaccess.com> wrote:

Hey all;

I'm planning on using a MAX4622 CMOS switch in a circuit I'm building. On
p.7 of the data sheet
(http://datasheets.maxim-ic.com/en/ds/MAX4621-MAX4623.pdf), it says:

"If power-supply sequencing is not possible, add two small signal diodes
in
series with the supply pins and a Schottky diode between V+ and VL"

I've read a bit about Schottky diodes, and I see that there are several
different types (like with normal diodes I suppose). How do I know what I
should use for this purpose? Can someone recommend one?

Thanks



The need for extra diodes suggests that this part has SCR latchup
problems. If it does, other things, like signal spikes, could also
latch it up... and then it will fry.

Try to find somebody else's part. Maxim tends to be bad news, in
general, anyhow.

Why do you need a 5 ohm analog switch?

John

I'm using it for switching different guitar effects pedals in and out of my
signal chain. I suppose I don't need 5 ohms, but I have a couple here and
thought I might as well use them, rather than order other parts.

Thanks
 
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

tempus fugit wrote:
Sorry, I forgot one other question. The datasheet says to sequence V+, then
Vl, V-, then logic ins. If the V+ was 9v, and Vl was 5v, and I had a power
supply that went:

+15vDC =>7809 => 7805, would that be power supply sequencing?
No. power supply sequencing is when the different power rails are
switched on during powerup with a delay between. Some ICs can become
unstable if all of the rails are applied at once. If you select
regulators with an enable pin (7809, 7805 do not) you can use a simple RC
circuit to delay powerup.

- --
Brendan Gillatt | GPG Key: 0xBF6A0D94
brendan {a} brendangillatt (dot) co (dot) uk
http://www.brendangillatt.co.uk
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (MingW32)

iD8DBQFKME1Erlc7Stqch40RAlFsAJ9IDosbGzBZbTccc4LuT17S9knpMQCeM3cC
soUBw+sCdNprjjj1WfC+lBk=
=IdiK
-----END PGP SIGNATURE-----
 
"Brendan Gillatt" <brendanREMOVETHIS@brendanREMOVETHISgillatt.co.uk> wrote
in message news:gvudnRWq6Ivf0K3XnZ2dnUVZ8q9i4p2d@pipex.net...
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

tempus fugit wrote:

Sorry, I forgot one other question. The datasheet says to sequence V+,
then
Vl, V-, then logic ins. If the V+ was 9v, and Vl was 5v, and I had a
power
supply that went:

+15vDC =>7809 => 7805, would that be power supply sequencing?

No. power supply sequencing is when the different power rails are
switched on during powerup with a delay between.
If the 7809 and 7805 are in series (IOW, output of 7809 feeds input of
7805), wouldn't there be a delay in between?

Thanks
 
tempus fugit wrote:
"Brendan Gillatt" <brendanREMOVETHIS@brendanREMOVETHISgillatt.co.uk> wrote
in message news:gvudnRWq6Ivf0K3XnZ2dnUVZ8q9i4p2d@pipex.net...

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

tempus fugit wrote:

Sorry, I forgot one other question. The datasheet says to sequence V+,

then

Vl, V-, then logic ins. If the V+ was 9v, and Vl was 5v, and I had a

power

supply that went:

+15vDC =>7809 => 7805, would that be power supply sequencing?

No. power supply sequencing is when the different power rails are
switched on during powerup with a delay between.


If the 7809 and 7805 are in series (IOW, output of 7809 feeds input of
7805), wouldn't there be a delay in between?

Thanks
Yes, but the 5V could become available to the circuit
_before_ the 9V output cap is fully charged.

Ed
 
Hi Tempus

This is what is happening...

The FIG in the Maxim datasheet shows internal diodes from the INPUTS to the
V+ line. If your input voltages are brought up before the V+ line, these
diodes will conduct, leak current into the chip and the chip may latch up.

What the diagram does not show is that there is probably a similar diode
between VL and V+. If you bring up the VL line before V+, then this diode
will, like the others, conduct.

You need to bypass this path with a diode with a lower conduction voltage.
The internal diodes are normal silicon diodes with a drop of 0.6V. A
Schottky diode has a drop of about 0.3V, so this will conduct before the
internal diodes.

The 2 diodes in series with the supply pin stop the voltage from the VL line
getting back into your V+ supply. When the V+ line comes up, these series
diodes will conduct and take over the powering of the chip from the (VL to
V+) diodes.

In the power supply configuration you are suggesting, all will be OK. The V+
line (15V) will come up first, the 7905 will take some time, but will come
up second then the 7805 will follow.

You then need to just supply V-

Hope this helps

Regards

--
Bill Naylor
www.electronworks.co.uk
Electronic Kits for Education and Fun


"tempus fugit" <toccata@quitspammingme.ciaccess.com> wrote in message
news:178af$4a304df6$d1d89dfd$7531@PRIMUS.CA...
"Brendan Gillatt" <brendanREMOVETHIS@brendanREMOVETHISgillatt.co.uk> wrote
in message news:gvudnRWq6Ivf0K3XnZ2dnUVZ8q9i4p2d@pipex.net...
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

tempus fugit wrote:

Sorry, I forgot one other question. The datasheet says to sequence V+,
then
Vl, V-, then logic ins. If the V+ was 9v, and Vl was 5v, and I had a
power
supply that went:

+15vDC =>7809 => 7805, would that be power supply sequencing?

No. power supply sequencing is when the different power rails are
switched on during powerup with a delay between.

If the 7809 and 7805 are in series (IOW, output of 7809 feeds input of
7805), wouldn't there be a delay in between?

Thanks
 
"Electronworks.co.uk" <newsgroups@electronworks.co.uk> wrote in message
news:xvSdnSiVaJzUW63XnZ2dnUVZ8kudnZ2d@bt.com...
Hi Tempus

This is what is happening...

The FIG in the Maxim datasheet shows internal diodes from the INPUTS to
the
V+ line. If your input voltages are brought up before the V+ line, these
diodes will conduct, leak current into the chip and the chip may latch up.

What the diagram does not show is that there is probably a similar diode
between VL and V+. If you bring up the VL line before V+, then this diode
will, like the others, conduct.

You need to bypass this path with a diode with a lower conduction voltage.
The internal diodes are normal silicon diodes with a drop of 0.6V. A
Schottky diode has a drop of about 0.3V, so this will conduct before the
internal diodes.

The 2 diodes in series with the supply pin stop the voltage from the VL
line
getting back into your V+ supply. When the V+ line comes up, these series
diodes will conduct and take over the powering of the chip from the (VL to
V+) diodes.

In the power supply configuration you are suggesting, all will be OK. The
V+
line (15V) will come up first, the 7905 will take some time, but will come
up second then the 7805 will follow.

You then need to just supply V-

Hope this helps

Regards

--
Bill Naylor
www.electronworks.co.uk
Electronic Kits for Education and Fun

Thanks for that Bill - that's much clearer now. Could you clarify something
though:

In the power supply configuration you are suggesting, all will be OK. The
V+
line (15V) will come up first, the 7905 will take some time, but will come
up second then the 7805 will follow.
I assume you mean 7809, not 7905 here?

So I don't need to worry about the protection diodes? Also, since I'm using
this device to switch audio signals, I need as close to the supply voltage
as possible. Would the diodes in series with the supply pin reduce the
voltage swing by another 0.6v?

Thanks
 
Hi Tempus

Ah! you spotted my deliberate mistake. Yes I meant the 7809.

The diodes in series with the supply will limit the signal swing to 1.2V
less than the supply (0.6V down from the top rail and 0.6V up from the V-).
A doube whammy

Bill

--
Bill Naylor
www.electronworks.co.uk
Electronic Kits for Education and Fun


"tempus fugit" <toccata@quitspammingme.ciaccess.com> wrote in message
news:7c894$4a310190$d1d89dab$25139@PRIMUS.CA...
"Electronworks.co.uk" <newsgroups@electronworks.co.uk> wrote in message
news:xvSdnSiVaJzUW63XnZ2dnUVZ8kudnZ2d@bt.com...
Hi Tempus

This is what is happening...

The FIG in the Maxim datasheet shows internal diodes from the INPUTS to
the
V+ line. If your input voltages are brought up before the V+ line, these
diodes will conduct, leak current into the chip and the chip may latch
up.

What the diagram does not show is that there is probably a similar diode
between VL and V+. If you bring up the VL line before V+, then this diode
will, like the others, conduct.

You need to bypass this path with a diode with a lower conduction
voltage.
The internal diodes are normal silicon diodes with a drop of 0.6V. A
Schottky diode has a drop of about 0.3V, so this will conduct before the
internal diodes.

The 2 diodes in series with the supply pin stop the voltage from the VL
line
getting back into your V+ supply. When the V+ line comes up, these series
diodes will conduct and take over the powering of the chip from the (VL
to
V+) diodes.

In the power supply configuration you are suggesting, all will be OK. The
V+
line (15V) will come up first, the 7905 will take some time, but will
come
up second then the 7805 will follow.

You then need to just supply V-

Hope this helps

Regards

--
Bill Naylor
www.electronworks.co.uk
Electronic Kits for Education and Fun


Thanks for that Bill - that's much clearer now. Could you clarify
something
though:

In the power supply configuration you are suggesting, all will be OK. The
V+
line (15V) will come up first, the 7905 will take some time, but will
come
up second then the 7805 will follow.

I assume you mean 7809, not 7905 here?

So I don't need to worry about the protection diodes? Also, since I'm
using
this device to switch audio signals, I need as close to the supply voltage
as possible. Would the diodes in series with the supply pin reduce the
voltage swing by another 0.6v?

Thanks
 
Maxim have a free sample request service - no questions asked, but in
production are difficult to get hold of.

If you only want a few, they are a good company supplying good components.
If you are going into production, you had better order well in advance!

--
Bill Naylor
www.electronworks.co.uk
Electronic Kits for Education and Fun


"tempus fugit" <toccata@quitspammingme.ciaccess.com> wrote in message
news:d1456$4a303ee4$d1d89dfd$3597@PRIMUS.CA...
"John Larkin" <jjlarkin@highNOTlandTHIStechnologyPART.com> wrote in
message
news:fpc035lsgurn8km6kr5otu2cg806lgf71a@4ax.com...
On Wed, 10 Jun 2009 18:18:39 -0400, "tempus fugit"
toccata@quitspammingme.ciaccess.com> wrote:

Hey all;

I'm planning on using a MAX4622 CMOS switch in a circuit I'm building.
On
p.7 of the data sheet
(http://datasheets.maxim-ic.com/en/ds/MAX4621-MAX4623.pdf), it says:

"If power-supply sequencing is not possible, add two small signal diodes
in
series with the supply pins and a Schottky diode between V+ and VL"

I've read a bit about Schottky diodes, and I see that there are several
different types (like with normal diodes I suppose). How do I know what
I
should use for this purpose? Can someone recommend one?

Thanks



The need for extra diodes suggests that this part has SCR latchup
problems. If it does, other things, like signal spikes, could also
latch it up... and then it will fry.

Try to find somebody else's part. Maxim tends to be bad news, in
general, anyhow.

Why do you need a 5 ohm analog switch?

John


I'm using it for switching different guitar effects pedals in and out of
my
signal chain. I suppose I don't need 5 ohms, but I have a couple here and
thought I might as well use them, rather than order other parts.

Thanks
 

Welcome to EDABoard.com

Sponsor

Back
Top